N A
Monthly
SQL injection in Apartment Visitors Management System V1.1's login form allows remote unauthenticated attackers to bypass authentication and extract database contents via the username parameter. The vulnerability scores 9.4 CVSS with network attack vector and low complexity. Public exploit code exists (SSVC confirms POC status), making this immediately exploitable. EPSS data unavailable, but SSVC framework rates it as automatable with partial technical impact, indicating high practical risk for internet-exposed installations.
Privilege escalation in Doorman API gateway v0.1.0 and v1.0.2 allows authenticated users to elevate their role to high-privileged (non-admin) accounts by directly editing the 'role' field via the /platform/user/{username} endpoint. The vulnerability stems from missing authorization checks on self-service user updates - any valid login credential is sufficient to escalate privileges to roles like 'manager' or 'developer'. CVSS 9.9 (Critical) reflects the Changed scope and broad compromise potential. SSVC indicates proof-of-concept code exists but exploitation requires human interaction (not automatable), suggesting targeted rather than mass-exploitation risk. No CISA KEV listing at time of analysis.
Local privilege escalation in DeepCool DeepCreative software version 1.2.7 and earlier allows unauthenticated attackers to execute arbitrary code with elevated privileges through malicious file processing. The vulnerability stems from insecure permission configuration (CWE-277) requiring user interaction to open a crafted file. Public exploit research exists on GitHub (uncle-hash repository), though CISA has not confirmed active exploitation. CVSS 7.8 indicates high severity, but EPSS data unavailable; SSVC framework rates technical impact as total with no confirmed exploitation and non-automatable attack path.
SQL injection in CodeAstro Simple Attendance Management System v1.0 allows remote unauthenticated attackers to bypass authentication via the username parameter in index.php. CVSS 9.8 (AV:N/AC:L/PR:N/UI:N) indicates trivial exploitation. CISA SSVC framework confirms proof-of-concept exists, attack is automatable, and technical impact is total (full system compromise). Public POC available on GitHub enables immediate weaponization by attackers with no specialized skills.
Server-Side Request Forgery (SSRF) in Craftql PHP library versions 1.3.7 and earlier enables remote attackers to force the server to make unintended requests, potentially leading to arbitrary code execution. The vulnerability resides in the GetAssetsFieldSchema.php listener component. No active exploitation is confirmed (not in CISA KEV), but a proof-of-concept repository with detailed exploitation documentation exists on GitHub. Despite the CVSS 7.5 rating, the extremely low EPSS score (0.01%, 0th percentile) indicates minimal real-world exploitation activity observed to date. The description claims RCE capability, but the CVSS vector shows only confidentiality impact (C:H/I:N/A:N), suggesting the SSRF may enable information disclosure that could chain into RCE rather than direct code execution - verification with vendor advisories needed.
Password reset token leakage in FuelCMS 1.5.2 allows unauthenticated attackers to steal a victim's password reset token through a malicious link placed in a valid password reset email. The attacker can then reset the victim's password, gaining unauthorized account access. A proof-of-concept exploit is publicly available, though exploitation is not automated and no active exploitation has been confirmed.
Remote code execution in Slah CMS v1.5.0 and below allows unauthenticated attackers to inject and execute arbitrary code via a crafted input to the session() function in config.php. Proof-of-concept exploit code is publicly available, though no active exploitation has been confirmed by CISA KEV. With a CVSS score of 9.8 and a full-impact compromise, this vulnerability poses a critical risk to any exposed instance.
SQL injection in Slah CMS v1.5.0 and below permits unauthenticated remote attackers to execute arbitrary SQL via the id parameter in vereador_ver.php, risking data breach and tampering. No active exploitation or public exploit is known, and EPSS indicates a 0.03% probability of exploitation.
Sensitive information disclosure in Slah v1.5.0 and below allows unauthenticated remote attackers to read the contents of config.php and extract active session credentials. The vulnerability permits high-confidentiality impact with no user interaction required, enabling session hijacking and unauthorized access. A proof-of-concept exploit is publicly available,
Directory traversal in SAC-NFe v2.0.02 allows unauthenticated attackers to read arbitrary files from the server via crafted GET requests to download.php. The flaw exposes sensitive files, potentially including credentials or configuration data, and a proof-of-concept exploit is publicly available, though active exploitation has not been confirmed.
Authenticated remote code execution in Daylight Studio FuelCMS v1.5.2 allows low-privileged attackers to execute arbitrary operating system commands via the add_git_submodule method in the Installer controller. A publicly available proof-of-concept exploit exists, but no active exploitation has been reported and EPSS likelihood remains very low at 0.05%.
Remote code execution in Upsonic 0.71.6 allows unauthenticated attackers to execute arbitrary OS commands via crafted MCP tasks. Although an allowlist restricts commands, the permitted npm and npx commands accept flags that enable OS command injection. The vulnerability is automatable with total technical impact, but no exploitation has been observed and the EPSS score is very low (0.04%).
Remote code execution in Agent Zero 0.9.8 via unsanitized command execution in its External MCP Servers feature allows unauthenticated attackers to inject arbitrary operating system commands. By crafting a malicious MCP server definition with attacker-controlled command and args, a remote attacker can achieve code execution with the privileges of the Agent Zero process, potentially leading to full system compromise. No active exploitation or public exploit is currently known, and EPSS indicates low exploitation probability.
Remote code execution in LangChain-ChatChat 0.3.1 allows unauthenticated remote attackers to execute arbitrary commands by injecting malicious commands into the MCP STDIO server configuration via the exposed management interface. When the MCP feature is enabled for agent execution, subsequent agent activity triggers the configured command, leading to full code execution in the service context. No active exploitation has been reported, and EPSS indicates a low probability of widespread attack.
Remote command injection in Jaaz 1.0.30 allows unauthenticated attackers to execute arbitrary commands on the underlying server via crafted MCP STDIO requests. The vulnerability resides in the application's handling of the Model Context Protocol (MCP) over standard input/output, and despite a CVSS score of 7.3 with low impact metrics, the description indicates full compromise is possible. No active exploitation or public exploit code has been identified, but the flaw is automatable and remotely exploitable without privileges.
Remote command execution via prompt injection in Windsurf 1.9544.26 allows attackers to register malicious MCP STDIO servers through crafted HTML content, leading to arbitrary command execution without further user interaction after initial payload delivery. No active exploitation or public exploit code has been reported, and EPSS suggests very low exploitation probability (0.04%). The vulnerability has a CVSS base score of 8.0, but the local attack vector and required user interaction significantly limit real-world risk.
CentSDR's Thread1 function suffers a remote stack overflow, leading to denial of service without authentication. A public proof-of-concept is available, but no active exploitation has been observed. The low EPSS score and niche user base suggest limited real-world risk despite the high CVSS score.
Broken Object-Level Authorization in Webkul Krayin CRM v2.2.x allows authenticated attackers to access, modify, and delete contact records belonging to other users without authorization. The vulnerability exists in the PersonController.php endpoint where insufficient access controls permit low-privileged authenticated users to manipulate arbitrary contact objects via crafted GET requests. No public exploit identified at time of analysis, though the attack vector is network-accessible with low complexity (CVSS:3.1 AV:N/AC:L/PR:L).
Broken Object-Level Authorization in Webkul Krayin CRM 2.2.x allows authenticated attackers to read, modify, and permanently delete leads belonging to other users via crafted GET requests to the LeadController endpoint. This multi-tenant access control failure enables horizontal privilege escalation across customer boundaries. Authentication requirements are minimal (PR:L in CVSS vector), and the attack complexity is low with no user interaction required. No public exploit identified at time of analysis, though the vulnerability appears straightforward to exploit given its BOLA nature.
Authenticated attackers can reset arbitrary user passwords in Webkul Krayin CRM v2.2.x through a Broken Object-Level Authorization (BOLA) vulnerability in the /Settings/UserController.php endpoint, enabling full account takeover of any user account. The attack requires low-privilege authentication (PR:L) and is exploitable remotely with low complexity (AV:N/AC:L), presenting an 8.8 CVSS severity with high impact to confidentiality, integrity, and availability. EPSS data not available; no CISA KEV listing identified; publicly available exploit code exists per researcher disclosure.
SQL injection in Krayin CRM 2.2.x allows authenticated remote attackers to extract sensitive database contents via the rotten_lead parameter in LeadDataGrid.php. CVSS 7.1 severity with network attack vector and low complexity enables database enumeration with low-privilege credentials. No public exploit identified at time of analysis, though EPSS data unavailable. Technical advisory published on GitHub indicates vulnerability affects lead management functionality in this Laravel-based open-source CRM platform.
Server-Side Request Forgery in Webkul Krayin CRM 2.2.x enables authenticated users to scan internal network resources and access sensitive information through the webhook creation endpoint. Attackers with low-privilege accounts can send crafted POST requests to /settings/webhooks/create, forcing the server to make requests to arbitrary internal URLs. With CVSS 8.5 (High) and scope change to other components, this allows reconnaissance of internal infrastructure, access to cloud metadata endpoints, and potential lateral movement. EPSS data not available; no public exploit identified at time of analysis, though technical details are published in security advisory.
Remote code execution in Webkul Krayin CRM v2.2.x allows authenticated attackers with low-level privileges to upload and execute malicious PHP files through an unrestricted file upload vulnerability in the /admin/tinymce/upload endpoint. The scope change (CVSS S:C) indicates potential container escape or cross-tenant impact. Authentication requirement confirmed (CVSS PR:L). No public exploit identified at time of analysis, though technical details are available in security advisory references.
Formula/CSV injection in HostBill billing automation builds 2025-11-24 and 2025-12-01 lets a remote unauthenticated attacker plant malicious spreadsheet formulas through a user-facing registration field, which are later exported and detonated when an operator opens the generated CSV, yielding code execution on the operator's workstation and potential privilege escalation. Tracked as CWE-1236 with a 9.8 CVSS and SSVC classifying it as PoC-available, automatable, with total technical impact; publicly available exploit code exists via a third-party GitHub write-up, but it is not on CISA KEV and EPSS is low at 0.10% (28th percentile). The vendor has published a security advisory and release notes covering the affected builds.
Remote code execution in Pyro (Python Remote Objects) 3.x arises because the wire protocol deserializes untrusted client messages with Python's pickle module, letting an attacker who can reach a Pyro daemon run arbitrary code by sending a crafted pickled message. All Pyro 3.x deployments that accept network connections are affected, and CWE-94 code injection combined with the CVSS 9.8 rating reflects full compromise of the host process. There is no public exploit identified at time of analysis, and the low EPSS score (0.02%) suggests no observed opportunistic exploitation despite the trivial technical requirements.
Command injection in KubePlus 4.14's kubeconfiggenerator component allows an attacker to run arbitrary shell commands on the host by supplying a crafted `chartName` value to the `/registercrd` endpoint. The flaw stems from unsanitized concatenation into a `subprocess.Popen(..., shell=True)` call, yielding full code execution (SSVC technical impact: total). A public proof-of-concept exists, but EPSS is very low (0.04%) and it is not on CISA KEV, so no public exploit indicates active exploitation at time of analysis.
SQL injection in Sourcecodester Online Thesis Archiving System v1.0's /otas/view_archive.php endpoint allows remote unauthenticated attackers to manipulate database queries, potentially extracting sensitive thesis data, authentication credentials, or modifying database contents. No public exploit identified at time of analysis, with minimal observed exploitation probability (EPSS 0.01%, 2nd percentile). The vulnerability affects a PHP-based academic archiving platform with limited deployment footprint.
Heap buffer overflow in FFmpeg 8.0.1's av_bprint_finalize() function enables remote denial-of-service attacks through maliciously crafted media files. Exploitation requires no authentication (CVSS AV:N/AC:L/PR:N/UI:N), making this accessible to any network-based attacker who can deliver manipulated input to vulnerable FFmpeg instances. EPSS score of 0.04% (12th percentile) indicates low observed exploitation probability, and no public exploit code or active exploitation has been identified at time of analysis. While CVSS rates this 7.5 HIGH due to availability impact, real-world risk is primarily limited to public-facing media processing services.
Resource deallocation flaw in FFmpeg 8.0.1's zmqsend.c utility enables remote denial of service through crafted input files. Attackers can trigger improper cleanup of allocated resources without authentication (CVSS AV:N/AC:L/PR:N/UI:N), exhausting system resources. EPSS score of 0.04% (11th percentile) indicates low real-world exploitation probability, and no active exploitation confirmed (not in CISA KEV). The vulnerability affects a non-core utility component used for ZeroMQ message sending, limiting practical attack surface compared to main FFmpeg libraries.
Out-of-bounds read in FFmpeg 8.0.1's AV1 decoder allows remote denial-of-service via malicious video files. Attackers craft inputs targeting read_global_param() in libavcodec/av1dec.c to trigger memory access violations, crashing the decoder. Affects applications processing untrusted AV1 video content (media servers, transcoders, browsers with FFmpeg). CVSS 7.5 (High) reflects network-exploitable DoS; EPSS 0.04% indicates low observed exploitation probability. No active exploitation confirmed (n
Totara LMS versions up to 19.1.5 allow unauthenticated remote attackers to conduct email bombing attacks by abusing the forgot password API endpoint, which lacks rate limiting on target email addresses. With a CVSS score of 9.8, the vulnerability enables complete compromise of confidentiality, integrity, and availability. Despite the critical score, EPSS estimates only 0.02% exploitation probability (5th percentile), and no active exploitation is confirmed (not in CISA KEV). A public proof-of-concept exists on GitHub, demonstrating the abuse technique.
Brute-force authentication bypass in Totara LMS versions 19.1.5 and earlier allows unauthenticated remote attackers to compromise user accounts via credential stuffing. The vulnerability chains login page manipulation with missing rate-limiting controls, enabling automated password guessing attacks. With CVSS 9.8 (critical) severity but only 0.02% EPSS probability (5th percentile), this represents a high-severity design flaw with currently low observed exploitation activity. A proof-of-concept exploit exists on GitHub (saykino/CVE-2026-31282), lowering the barrier for attack automation, though no confirmed active exploitation (CISA KEV) has been reported.
HTML Injection in Totara LMS through version 19.1.5 allows authenticated users with low privileges to inject malicious HTML/JavaScript into messages sent to all application users, enabling session hijacking and arbitrary command execution in victims' browsers. A publicly available exploit exists (GitHub POC referenced), though no confirmed active exploitation (not in CISA KEV). EPSS score of 0.02% indicates very low observed exploitation probability despite the CVSS 8.0 rating, suggesting limited attacker interest or opportunity in real-world environments.
Hardcoded debug credentials in BMC Control-M/MFT 9.0.20-9.0.22 enable unauthenticated remote attackers to access the MFT API debug interface with full system privileges. The vulnerability (CWE-798) stems from cleartext default credentials embedded in the application package, providing complete confidentiality, integrity, and availability compromise (CVSS 9.8). No public exploit identified at time of analysis, with EPSS score of 0.02% indicating low observed exploitation activity despite critical severity rating.
SQL injection in SourceCodester Engineers Online Portal v1.0 allows remote unauthenticated attackers to execute arbitrary SQL commands through the new_password parameter in update_password.php, enabling complete database compromise, authentication bypass, and potential server takeover. Despite a critical CVSS score of 9.8, EPSS indicates only 0.01% exploitation probability (2nd percentile), and no public exploit identified at time of analysis beyond technical documentation. The vulnerable PHP application lacks input validation on password update functionality, a common weakness in legacy web portals.
SQL injection in itsourcecode Online Student Enrollment System v1.0 allows unauthenticated remote attackers to execute arbitrary SQL commands via the 'subjcode' parameter in assignInstructorSubjects.php, potentially enabling complete database compromise, authentication bypass, and data exfiltration. CVSS 9.8 (Critical) reflects network-accessible exploitation requiring no privileges or user interaction. EPSS score of 0.01% (2nd percentile) indicates minimal observed exploitation activity, though publicly available exploit code exists via GitHub repository documentation. No vendor-released patch identified at time of analysis.
SQL injection in itsourcecode Online Student Enrollment System v1.0 allows unauthenticated remote attackers to extract, modify, or delete database contents via unsanitized classId parameter in instructorClasses.php. Despite critical CVSS 9.8 rating, EPSS score of 0.01% (2nd percentile) indicates minimal observed exploitation activity. Public proof-of-concept documentation exists on GitHub, lowering exploitation barrier for opportunistic attackers targeting educational institutions using this enrollment platform.
SQL injection in PHP-MYSQL-User-Login-System v1.0 allows unauthenticated remote attackers to execute arbitrary SQL commands via the username parameter at login.php, potentially enabling complete database compromise, authentication bypass, and unauthorized administrative access. Publicly available exploit code exists (GitHub POC). Despite a critical CVSS score of 9.8, EPSS exploitation probability is extremely low (0.01%, 2nd percentile), and no confirmed active exploitation (not in CISA KEV). The low real-world exploitation signal suggests limited deployment or attacker interest in this specific application.
SQL injection in itsourcecode Online Student Enrollment System v1.0 allows unauthenticated remote attackers to execute arbitrary SQL commands through the 'subjcode' parameter in scheduleSubList.php, achieving full database compromise with potential for complete system control. CVSS 9.8 (Critical) reflects network-accessible exploitation requiring no privileges or user interaction. EPSS score of 0.01% (2nd percentile) indicates minimal observed exploitation activity, and no CISA KEV listing exists, suggesting limited real-world targeting despite the theoretical severity. Public proof-of-concept documentation exists on GitHub, lowering the technical barrier for exploitation.
SQL injection in itsourcecode Online Student Enrollment System v1.0 allows unauthenticated remote attackers to execute arbitrary SQL commands via the 'coursename' parameter in newCourse.php, potentially enabling complete database compromise, authentication bypass, and server-side code execution. Despite a critical 9.8 CVSS score, EPSS indicates only 0.01% exploitation probability (2nd percentile), suggesting minimal observed attacker interest. Publicly available exploit code exists (GitHub POC), but no CISA KEV listing confirms active exploitation.
BMC Control-M/MFT 9.0.20-9.0.22 exposes API identifiers and secret values to unauthenticated remote attackers via an API management endpoint, enabling unauthorized invocation of privileged API operations. The CVSS vector (AV:N/AC:L/PR:N) confirms network-accessible, low-complexity exploitation requiring no authentication. Despite a 7.5 CVSS score, EPSS probability is minimal (0.03%, 8th percentile), and no public exploit or active exploitation (non-KEV) is identified at time of analysis.
SQL injection in BMC Control-M/MFT 9.0.20-9.0.22 debug API enables authenticated attackers to execute arbitrary SQL queries, leading to file system access and potential remote code execution. CVSS 8.8 reflects network-accessible, low-complexity exploitation requiring low-privilege authentication. EPSS score of 0.04% (13th percentile) indicates minimal current exploitation probability, and no public exploit identified at time of analysis. Patch PAAFP-9-0-22-025 addresses the vulnerability.
PowerStrip driver (pstrip64.sys) through version 3.90.736 enables authenticated local users to escalate privileges to SYSTEM by sending malicious IOCTL requests that map arbitrary physical memory into user-mode address space, allowing modification of kernel structures. EPSS score of 0.02% (5th percentile) indicates low automated exploitation likelihood, though publicly available exploit code exists (PacketStorm Security reference), making this a realistic threat in environments where PowerStrip is deployed and local access is possible. No CISA KEV listing indicates no confirmed widespread exploitation at time of analysis.
Command injection in ToToLink A3300R router firmware v17.0.0cu.557_B20221024 enables unauthenticated remote attackers to execute arbitrary OS commands via the stun-pass parameter in /cgi-bin/cstecgi.cgi endpoint. No public exploit identified at time of analysis, despite GitHub vulnerability disclosure repository. CVSS 9.8 (Critical) reflects trivial network-based exploitation without authentication, though EPSS probability remains low (0.01%, 2nd percentile), suggesting limited current attacker interest in this specific router model despite maximum severity rating.
DLL injection in GatewayGeo MapServer for Windows version 5 enables authenticated local attackers to escalate privileges to SYSTEM level through crafted executable placement. The vulnerability exploits insecure library loading paths, allowing low-privileged users to inject malicious DLLs that execute with elevated permissions. Publicly available exploit code exists. Affects Windows deployments only; CVSS 8.8 reflects local attack vector requiring low privileges but achieving full system compromise across security boundaries.
Remote code execution in OSGeo MapServer versions prior to 8.0 enables unauthenticated attackers to execute arbitrary code through dynamic-link library (DLL) injection via a specially crafted executable. The vulnerability requires no user interaction and has low attack complexity (CVSS 9.1 Critical), though real-world exploitation probability remains low (EPSS 2%, 5th percentile). Publicly available exploit code exists in a researcher's GitHub repository, but no confirmed active exploitation (CISA KEV) has been documented at time of analysis.
Server-Side Request Forgery in ERPNext 16.0.1 and Frappe Framework 16.1.1 enables unauthenticated attackers to force servers to make arbitrary HTTP requests to internal services through insufficiently sanitized HTML in Print Format PDF generation. Attackers inject HTML elements like <iframe> referencing external resources, which the PDF rendering engine automatically fetches server-side, exposing cloud metadata endpoints and internal network resources. No public exploit identified at time of analysis. CVSS 9.1 severity reflects network-accessible attack vector requiring no authentication or user interaction.
Buffer overflow in OpenAirInterface 2.2.0 AUSF component crashes service when processing oversized NAS PDU Authentication Response via UplinkNASTransport messages. Unauthenticated remote attackers can send malformed authentication responses (e.g., 100-byte payloads exceeding expected bounds) triggering AUSF component crash, preventing legitimate user registration and verification. Affects 5G core network deployments using OpenAirInterface AUSF. No public exploit identified at time of analysis. CVSS 7.5 High severity due to network-accessible denial of service without authentication requirements.
Integrity protection bypass in OpenAirInterface v2.2.0 allows unauthenticated network attackers to downgrade 5G security context by forcing acceptance of IA0-only capability during initial UE registration, despite NIA1/NIA2 being configured. Exploitation enables replay attacks against mobile network infrastructure through manipulation of Security Mode Complete messages, compromising session integrity without confidentiality impact. No public exploit identified at time of analysis.
Unauthenticated super administrator account creation in MRCMS 3.1.2 allows remote attackers to bypass all access controls and add privileged accounts directly via UserController.save() method. The vulnerability exposes full system compromise through network-accessible endpoints requiring no prior authentication. CVSS 9.8 critical severity reflects unrestricted administrative takeover. No public exploit identified at time of analysis; low observed exploitation activity (EPSS <1%).
Unauthenticated account creation bypass in megagao production_ssm v1.0 allows remote attackers to create super administrator accounts via direct API access to /user/insert endpoint. The UserController.java insert() method processes account creation requests without authentication enforcement (CVSS vector PR:N confirms unauthenticated access). Successful exploitation grants full administrative control, enabling attackers to compromise confidentiality, integrity, and availability of the entire application. No public exploit identified at time of analysis.
Authenticated remote code execution in Daylight Studio FuelCMS version 1.5.2 allows low-privileged users to execute arbitrary code via the Blocks module. CVSS 8.8 rating indicates network-accessible attack requiring low-complexity exploitation without user interaction, enabling full system compromise (confidentiality, integrity, availability impact). No public exploit identified at time of analysis. Low observed exploitation activity (EPSS 0.02%).
Authentication bypass in OpenAirInterface V2.2.0 Access Management Function (AMF) allows unauthenticated remote attackers to register unauthorized User Equipment (UE) devices on 5G core networks. Exploiting incorrect state machine transitions during UE registration, attackers send SecurityModeComplete messages after InitialUERegistration to trigger registration acceptance without completing proper authentication procedures. This grants full network access to malicious devices, enabling unauthorized subscriber services consumption, interception of traffic, and potential lateral movement within 5G infrastructure. No public exploit identified at time of analysis.
Authentication bypass in Kaleris Yard Management System (YMS) v7.2.2.1 enables unauthenticated remote attackers to completely circumvent login verification and gain unauthorized access to application resources with full confidentiality, integrity, and availability impact. The vulnerability has a 9.8 CVSS score with network-based attack vector requiring no privileges or user interaction. Currently tracked at 2% EPSS (5th percentile) with no confirmed active exploitation (not in CISA KEV), though a public proof-of-concept repository exists on GitHub, significantly elevating exploitation risk for this critical authentication flaw.
AMF daemon in OpenAirInterface V2.2.0 crashes upon receipt of malformed NGAP control-plane messages containing mismatched procedure codes or PDU type violations (e.g., successfulOutcome where InitiatingMessage is required). Remote attackers can trigger denial of service without authentication (CVSS AV:N/PR:N), exploiting improper input validation (CWE-20) in 5G core network signaling. Publicly available exploit code exists, SSVC framework classifies as automatable with partial technical impact, and EPSS data not provided but attack simplicity (AC:L) indicates low barrier to exploitation.
Percona PMM (Percona Monitoring and Management) versions prior to 3.7 allow authenticated users with pmm-admin privileges to execute arbitrary operating system commands by exploiting excessive database superuser privileges through the 'Add data source' feature. This privilege escalation vulnerability enables container/system breakout from database context to shell access. EPSS score is low (0.04%, 13th percentile) indicating minimal observed exploitation activity, and CISA SSVC assessment confirms no active exploitation with non-automatable attack characteristics, though technical impact is rated total.
Out-of-bounds read in Mbed TLS 3.x before 3.6.6 allows attackers to leak adjacent CCM context data through the multipart CCM API by passing an oversized tag_len parameter to mbedtls_ccm_finish(), which lacks validation against the internal 16-byte authentication buffer. Mbed TLS 4.x contains the same vulnerability in internal code but does not expose the vulnerable function publicly; exploitation requires direct application-level invocation of the affected API. No public exploit code or active exploitation has been reported, but the attack requires no special privileges.
TOCTOU race condition in Balena Etcher for Windows (versions prior to 2.1.4) enables local privilege escalation to arbitrary code execution when attackers replace legitimate scripts with malicious payloads during disk flashing operations. The vulnerability requires low privileges and user interaction but achieves high impact across confidentiality, integrity, and availability with scope change. No public exploit identified at time of analysis, though technical details are available via researcher disclosure (B1tBreaker). EPSS data not available, but the local attack vector and high complexity reduce immediate remote exploitation risk.
Mbed TLS versions 3.5.0 through 4.0.0 allow client impersonation during TLS 1.3 session resumption, enabling an attacker to assume the identity of a legitimate client when reestablishing a previously negotiated session. The vulnerability affects the session resumption mechanism in TLS 1.3 and permits information disclosure; no CVSS score or exploit status data is currently available from public sources.
Finite-field Diffie-Hellman (FFDH) in Mbed TLS 3.5.x, 3.6.0 through 3.6.5, and TF-PSA-Crypto 1.0 lacks contributory behavior due to improper validation of peer-supplied parameters, allowing an attacker to restrict the shared secret to a small set of predictable values. While the vulnerability does not directly impact TLS (which does not depend on contributory behavior), it poses a significant risk to protocols that do rely on this property, including those where an active network attacker or malicious peer can exploit the weakness. No CVSS score or public exploit code has been assigned at the time of analysis.
NULL pointer dereference in Mbed TLS distinguished name (X.509) parsing allows remote attackers to trigger a denial of service by writing to address 0, affecting Mbed TLS versions 3.6.5 and earlier, and 4.0.0. The vulnerability is reachable during X.509 certificate processing and does not require authentication. No public exploit code or active exploitation has been confirmed at the time of analysis.
Mbed TLS before version 3.6.6 and TF-PSA-Crypto before version 1.1.0 contain a PRNG seed misuse vulnerability that enables information disclosure. An attacker who gains access to a seeded PRNG instance can potentially predict or replicate pseudo-random number generation, compromising cryptographic material confidentiality. The vulnerability affects cryptographic libraries used in embedded systems and IoT devices, with confirmed availability of vendor security advisories but no CVSS score assigned at time of analysis.
Buffer overflow in Mbed TLS versions 3.5.0 through 3.6.5 allows remote attackers to cause a denial of service or potentially execute arbitrary code via crafted input to the x509_inet_pton_ipv6() function used in X.509 certificate parsing. The vulnerability is fixed in Mbed TLS 3.6.6 and 4.1.0. No public exploit code or confirmed active exploitation has been identified at the time of analysis.
Buffer overflow in Mbed TLS public key export functionality for Finite Field Diffie-Hellman (FFDH) keys affects versions through 3.6.5 and TF-PSA-Crypto 1.0.0. An attacker can trigger a memory corruption condition during FFDH public key export operations, potentially enabling code execution or denial of service depending on memory layout and application context. No public exploit code or active exploitation has been confirmed at time of analysis.
Remote code execution in DedeCMS 5.7.118 allows unauthenticated attackers to execute arbitrary code through crafted setup tag values during module upload operations. The vulnerability exploits insufficient input validation in the module upload functionality, enabling direct code injection. No CVSS score, EPSS data, or KEV confirmation is available; however, the presence of a public proof-of-concept demonstrates practical exploitability.
SQL injection in pandas-ai v3.0.0 allows remote code execution through the pandasai.agent.base._execute_sql_query component, enabling attackers to manipulate SQL queries and potentially access, modify, or exfiltrate database contents. No CVSS score, EPSS data, or KEV status is available; however, the vulnerability affects a widely-used data analysis library and publicly available proof-of-concept code exists, elevating real-world risk despite incomplete severity metrics.
Buffer overflow in TOTOlink A3600R v5.9c.4959 setAppEasyWizardConfig interface allows remote code execution or denial of service via unvalidated rootSsid parameter in /lib/cste_modules/app.so. The vulnerability affects a Wi-Fi router's configuration endpoint and enables unauthenticated attackers to trigger memory corruption with potential for arbitrary code execution. No CVSS vector or patch status was available at time of analysis.
Pharmacy Product Management System 1.0 accepts negative price and total cost values in sales transactions due to insufficient input validation in add-sales.php, enabling attackers to manipulate financial records, corrupt sales reports, and cause financial loss. The vulnerability allows unauthenticated or low-privilege users to submit arbitrary negative values that bypass business logic controls. Publicly available exploit code exists demonstrating this business logic flaw.
Arbitrary file overwrite in Docudepot PDF Reader v1.0.34 enables attackers to overwrite critical internal files through the file import process, resulting in arbitrary code execution or information exposure. The vulnerability affects the mobile PDF viewer application across Android platforms. No public exploit code or active exploitation has been confirmed at time of analysis, though the severity of potential impact (RCE) warrants immediate investigation and patching.
Arbitrary file overwrite in Ora Tools PDF Reader & Editor APP v4.3.5 enables attackers to overwrite critical internal files through the file import process, resulting in arbitrary code execution or information exposure. The vulnerability affects the Android application and has been publicly disclosed; however, CVSS scoring, CISA KEV status, and vendor patch availability have not been independently confirmed at time of analysis.
Arbitrary file overwrite in Zora: Post, Trade, Earn Crypto v2.60.0 enables attackers to overwrite critical internal files through the file import process, resulting in remote code execution or information exposure. The vulnerability affects the cryptocurrency trading application's file handling mechanism, allowing unauthenticated remote attackers to inject malicious content into system-critical files. No active exploitation has been confirmed at time of analysis, though the attack vector and impact severity warrant immediate investigation by affected users.
Arbitrary file overwrite in InTouch Contacts & Caller ID APP v6.38.1 allows remote attackers to overwrite critical internal files through the file import process, enabling arbitrary code execution or sensitive information exposure. Affected versions are limited to 6.38.1; no CVSS score, EPSS, or active exploitation status (KEV) is available at this time, though the vulnerability chain to RCE presents material risk.
Arbitrary file overwrite in PEAKSEL D.O.O. NIS Animal Sounds and Ringtones v1.3.0 allows attackers to overwrite critical internal files during the file import process, enabling remote code execution or sensitive information exposure. The vulnerability affects the application's import functionality without requiring authentication. No public exploit code or active exploitation has been independently confirmed at the time of analysis.
Arbitrary file overwrite in My Location Travel Timeline v11.80 by Squareapps LLC permits attackers to overwrite critical internal files through the file import process, resulting in arbitrary code execution or information disclosure. Attack vector and complexity details are not confirmed from available CVSS data, and active exploitation status is unconfirmed.
Arbitrary file overwrite in Funambol Zefiro Cloud v32.0.2026011614 allows attackers to overwrite critical internal files during the file import process, enabling remote code execution or information disclosure. The vulnerability affects the cloud application and its associated mobile client. No CVSS score or official vendor patch has been assigned as of analysis time, though the reported impact (RCE/information exposure) is severe.
Arbitrary file overwrite in UXGROUP LLC Cast to TV Screen Mirroring v2.2.77 enables remote attackers to overwrite critical application files through a malicious file import process, resulting in remote code execution or information disclosure. No CVSS score, exploit code availability, or active exploitation status confirmed from available data.
Arbitrary file overwrite in FLY is FUN Aviation Navigation v35.33 permits attackers to overwrite critical internal files through the file import process, enabling remote code execution or information disclosure. No CVSS score, CVE severity classification, or patch status has been established. The vulnerability affects a niche aviation navigation software product with limited public disclosure.
Arbitrary file overwrite in PDF Reader App TA/UTAX Mobile Print v3.7.2.251001 allows remote attackers to overwrite critical internal files during the file import process, potentially leading to remote code execution or unauthorized information exposure. The vulnerability affects a mobile print utility with demonstrated proof-of-concept documentation available on GitHub, though CVSS scoring and formal vendor patch status remain unavailable at time of analysis.
Arbitrary file overwrite in MaruNuri LLC v2.0.23 allows remote attackers to overwrite critical internal files during the file import process, enabling arbitrary code execution or information exposure. No CVSS score, exploit code availability, or active exploitation status is documented in available sources.
Arbitrary file overwrite in DeftPDF Document Translator v54.0 permits attackers to overwrite critical internal files through the file import mechanism, potentially enabling remote code execution or sensitive information exposure. The vulnerability affects DeftPDF Document Translator specifically at version 54.0 and is documented by academic researchers at Fudan University's security systems group. Attack complexity and authentication requirements cannot be definitively assessed due to missing CVSS vector data, though the file import process suggests user interaction may be required.
Remote code execution in Ridvay Code's command auto-approval module allows unauthenticated attackers to bypass whitelist security controls via shell command substitution syntax (e.g., $(...) or backticks) embedded in command arguments. The vulnerability stems from insufficient regular expression validation that fails to detect command injection payloads, permitting an attacker to execute arbitrary OS commands with automatic approval. No user interaction is required; a crafted command such as git log --grep="$(malicious_command)" will be misidentified as safe and executed by the underlying shell, resulting in remote code execution.
Directory traversal in agentic-context-engine up to version 0.7.1 enables arbitrary file writes through the checkpoint_dir parameter in OfflineACE.run, exploiting inadequate path normalization in the save_to_file method. Unauthenticated attackers can overwrite arbitrary files within the application process's permissions scope, potentially achieving code execution, privilege escalation, or application compromise depending on deployment context and file system layout.
Remote code execution in DSAI-Cline's command auto-approval module allows unauthenticated attackers to bypass whitelist validation by embedding literal newline characters within command payloads, forcing the system to execute arbitrary OS commands without user interaction. The vulnerability exploits ineffective string-based parsing that fails to sanitize newline separators, enabling attackers to chain whitelisted commands (e.g., git log) with malicious code that PowerShell interprets as sequential commands. No CVSS score, EPSS data, or KEV confirmation available; exploitation status and real-world impact remain unconfirmed.
Remote code execution in Ridvay Code's command auto-approval module allows unauthenticated attackers to bypass whitelist protections via shell command substitution syntax ($(…) and backticks) embedded in seemingly benign git commands, achieving code execution without user interaction. The vulnerability exploits inadequate regular expression validation that fails to detect shell metacharacters in command arguments, enabling attackers to inject arbitrary commands that execute with the privileges of the Ridvay Code process.
InfCode's terminal auto-execution module fails to properly validate PowerShell commands due to an ineffective blacklist and lack of semantic parsing, allowing attackers to bypass command filtering through syntax obfuscation. When a user imports a specially crafted file into the IDE, the Agent executes arbitrary PowerShell commands without user confirmation, leading to remote code execution or sensitive data exfiltration. No public exploit code or active exploitation has been confirmed at time of analysis.
Prompt injection attacks in Sixth's automatic terminal command execution feature bypass the model-based safety classification system, allowing attackers to execute arbitrary commands without user approval by wrapping malicious payloads in templates that mislead the AI into treating them as safe operations.
Remote code execution in DSAI-Cline's command auto-approval module allows unauthenticated attackers to bypass whitelist validation by embedding newline characters in command payloads, forcing automatic approval and sequential execution of arbitrary OS commands via PowerShell without user interaction.
HAI Build Code Generator's automatic command execution feature can be bypassed through prompt injection attacks, allowing unauthenticated remote code execution by misleading the AI model into misclassifying malicious commands as safe. The vulnerability exploits a fundamental design flaw in the model's safety classification logic, where attackers can wrap destructive commands in generic templates to bypass the user approval requirement that should be triggered for potentially dangerous operations.
SakaDev's automatic terminal command execution feature can be bypassed via prompt injection attacks, allowing unauthenticated remote attackers to execute arbitrary commands without user approval by wrapping malicious commands in templates that mislead the underlying language model into misclassifying destructive operations as safe. The vulnerability exploits a design flaw in the model-based safety classification mechanism rather than a traditional code defect, affecting the extension across all versions where the 'Execute safe commands' option is enabled.
Remote code execution in Syntx's command auto-approval module allows unauthenticated attackers to bypass whitelist security via shell command substitution syntax in command arguments. The vulnerability exploits inadequate regular expression parsing that fails to detect $(…) and backtick command substitution patterns, enabling an attacker to inject malicious commands within seemingly benign git operations (e.g., git log --grep="$(malicious_command)") that are automatically approved and executed with full system privileges. No CVSS score or KEV status data available; no public exploit code confirmed at time of analysis.
SQL injection in Apartment Visitors Management System V1.1's login form allows remote unauthenticated attackers to bypass authentication and extract database contents via the username parameter. The vulnerability scores 9.4 CVSS with network attack vector and low complexity. Public exploit code exists (SSVC confirms POC status), making this immediately exploitable. EPSS data unavailable, but SSVC framework rates it as automatable with partial technical impact, indicating high practical risk for internet-exposed installations.
Privilege escalation in Doorman API gateway v0.1.0 and v1.0.2 allows authenticated users to elevate their role to high-privileged (non-admin) accounts by directly editing the 'role' field via the /platform/user/{username} endpoint. The vulnerability stems from missing authorization checks on self-service user updates - any valid login credential is sufficient to escalate privileges to roles like 'manager' or 'developer'. CVSS 9.9 (Critical) reflects the Changed scope and broad compromise potential. SSVC indicates proof-of-concept code exists but exploitation requires human interaction (not automatable), suggesting targeted rather than mass-exploitation risk. No CISA KEV listing at time of analysis.
Local privilege escalation in DeepCool DeepCreative software version 1.2.7 and earlier allows unauthenticated attackers to execute arbitrary code with elevated privileges through malicious file processing. The vulnerability stems from insecure permission configuration (CWE-277) requiring user interaction to open a crafted file. Public exploit research exists on GitHub (uncle-hash repository), though CISA has not confirmed active exploitation. CVSS 7.8 indicates high severity, but EPSS data unavailable; SSVC framework rates technical impact as total with no confirmed exploitation and non-automatable attack path.
SQL injection in CodeAstro Simple Attendance Management System v1.0 allows remote unauthenticated attackers to bypass authentication via the username parameter in index.php. CVSS 9.8 (AV:N/AC:L/PR:N/UI:N) indicates trivial exploitation. CISA SSVC framework confirms proof-of-concept exists, attack is automatable, and technical impact is total (full system compromise). Public POC available on GitHub enables immediate weaponization by attackers with no specialized skills.
Server-Side Request Forgery (SSRF) in Craftql PHP library versions 1.3.7 and earlier enables remote attackers to force the server to make unintended requests, potentially leading to arbitrary code execution. The vulnerability resides in the GetAssetsFieldSchema.php listener component. No active exploitation is confirmed (not in CISA KEV), but a proof-of-concept repository with detailed exploitation documentation exists on GitHub. Despite the CVSS 7.5 rating, the extremely low EPSS score (0.01%, 0th percentile) indicates minimal real-world exploitation activity observed to date. The description claims RCE capability, but the CVSS vector shows only confidentiality impact (C:H/I:N/A:N), suggesting the SSRF may enable information disclosure that could chain into RCE rather than direct code execution - verification with vendor advisories needed.
Password reset token leakage in FuelCMS 1.5.2 allows unauthenticated attackers to steal a victim's password reset token through a malicious link placed in a valid password reset email. The attacker can then reset the victim's password, gaining unauthorized account access. A proof-of-concept exploit is publicly available, though exploitation is not automated and no active exploitation has been confirmed.
Remote code execution in Slah CMS v1.5.0 and below allows unauthenticated attackers to inject and execute arbitrary code via a crafted input to the session() function in config.php. Proof-of-concept exploit code is publicly available, though no active exploitation has been confirmed by CISA KEV. With a CVSS score of 9.8 and a full-impact compromise, this vulnerability poses a critical risk to any exposed instance.
SQL injection in Slah CMS v1.5.0 and below permits unauthenticated remote attackers to execute arbitrary SQL via the id parameter in vereador_ver.php, risking data breach and tampering. No active exploitation or public exploit is known, and EPSS indicates a 0.03% probability of exploitation.
Sensitive information disclosure in Slah v1.5.0 and below allows unauthenticated remote attackers to read the contents of config.php and extract active session credentials. The vulnerability permits high-confidentiality impact with no user interaction required, enabling session hijacking and unauthorized access. A proof-of-concept exploit is publicly available,
Directory traversal in SAC-NFe v2.0.02 allows unauthenticated attackers to read arbitrary files from the server via crafted GET requests to download.php. The flaw exposes sensitive files, potentially including credentials or configuration data, and a proof-of-concept exploit is publicly available, though active exploitation has not been confirmed.
Authenticated remote code execution in Daylight Studio FuelCMS v1.5.2 allows low-privileged attackers to execute arbitrary operating system commands via the add_git_submodule method in the Installer controller. A publicly available proof-of-concept exploit exists, but no active exploitation has been reported and EPSS likelihood remains very low at 0.05%.
Remote code execution in Upsonic 0.71.6 allows unauthenticated attackers to execute arbitrary OS commands via crafted MCP tasks. Although an allowlist restricts commands, the permitted npm and npx commands accept flags that enable OS command injection. The vulnerability is automatable with total technical impact, but no exploitation has been observed and the EPSS score is very low (0.04%).
Remote code execution in Agent Zero 0.9.8 via unsanitized command execution in its External MCP Servers feature allows unauthenticated attackers to inject arbitrary operating system commands. By crafting a malicious MCP server definition with attacker-controlled command and args, a remote attacker can achieve code execution with the privileges of the Agent Zero process, potentially leading to full system compromise. No active exploitation or public exploit is currently known, and EPSS indicates low exploitation probability.
Remote code execution in LangChain-ChatChat 0.3.1 allows unauthenticated remote attackers to execute arbitrary commands by injecting malicious commands into the MCP STDIO server configuration via the exposed management interface. When the MCP feature is enabled for agent execution, subsequent agent activity triggers the configured command, leading to full code execution in the service context. No active exploitation has been reported, and EPSS indicates a low probability of widespread attack.
Remote command injection in Jaaz 1.0.30 allows unauthenticated attackers to execute arbitrary commands on the underlying server via crafted MCP STDIO requests. The vulnerability resides in the application's handling of the Model Context Protocol (MCP) over standard input/output, and despite a CVSS score of 7.3 with low impact metrics, the description indicates full compromise is possible. No active exploitation or public exploit code has been identified, but the flaw is automatable and remotely exploitable without privileges.
Remote command execution via prompt injection in Windsurf 1.9544.26 allows attackers to register malicious MCP STDIO servers through crafted HTML content, leading to arbitrary command execution without further user interaction after initial payload delivery. No active exploitation or public exploit code has been reported, and EPSS suggests very low exploitation probability (0.04%). The vulnerability has a CVSS base score of 8.0, but the local attack vector and required user interaction significantly limit real-world risk.
CentSDR's Thread1 function suffers a remote stack overflow, leading to denial of service without authentication. A public proof-of-concept is available, but no active exploitation has been observed. The low EPSS score and niche user base suggest limited real-world risk despite the high CVSS score.
Broken Object-Level Authorization in Webkul Krayin CRM v2.2.x allows authenticated attackers to access, modify, and delete contact records belonging to other users without authorization. The vulnerability exists in the PersonController.php endpoint where insufficient access controls permit low-privileged authenticated users to manipulate arbitrary contact objects via crafted GET requests. No public exploit identified at time of analysis, though the attack vector is network-accessible with low complexity (CVSS:3.1 AV:N/AC:L/PR:L).
Broken Object-Level Authorization in Webkul Krayin CRM 2.2.x allows authenticated attackers to read, modify, and permanently delete leads belonging to other users via crafted GET requests to the LeadController endpoint. This multi-tenant access control failure enables horizontal privilege escalation across customer boundaries. Authentication requirements are minimal (PR:L in CVSS vector), and the attack complexity is low with no user interaction required. No public exploit identified at time of analysis, though the vulnerability appears straightforward to exploit given its BOLA nature.
Authenticated attackers can reset arbitrary user passwords in Webkul Krayin CRM v2.2.x through a Broken Object-Level Authorization (BOLA) vulnerability in the /Settings/UserController.php endpoint, enabling full account takeover of any user account. The attack requires low-privilege authentication (PR:L) and is exploitable remotely with low complexity (AV:N/AC:L), presenting an 8.8 CVSS severity with high impact to confidentiality, integrity, and availability. EPSS data not available; no CISA KEV listing identified; publicly available exploit code exists per researcher disclosure.
SQL injection in Krayin CRM 2.2.x allows authenticated remote attackers to extract sensitive database contents via the rotten_lead parameter in LeadDataGrid.php. CVSS 7.1 severity with network attack vector and low complexity enables database enumeration with low-privilege credentials. No public exploit identified at time of analysis, though EPSS data unavailable. Technical advisory published on GitHub indicates vulnerability affects lead management functionality in this Laravel-based open-source CRM platform.
Server-Side Request Forgery in Webkul Krayin CRM 2.2.x enables authenticated users to scan internal network resources and access sensitive information through the webhook creation endpoint. Attackers with low-privilege accounts can send crafted POST requests to /settings/webhooks/create, forcing the server to make requests to arbitrary internal URLs. With CVSS 8.5 (High) and scope change to other components, this allows reconnaissance of internal infrastructure, access to cloud metadata endpoints, and potential lateral movement. EPSS data not available; no public exploit identified at time of analysis, though technical details are published in security advisory.
Remote code execution in Webkul Krayin CRM v2.2.x allows authenticated attackers with low-level privileges to upload and execute malicious PHP files through an unrestricted file upload vulnerability in the /admin/tinymce/upload endpoint. The scope change (CVSS S:C) indicates potential container escape or cross-tenant impact. Authentication requirement confirmed (CVSS PR:L). No public exploit identified at time of analysis, though technical details are available in security advisory references.
Formula/CSV injection in HostBill billing automation builds 2025-11-24 and 2025-12-01 lets a remote unauthenticated attacker plant malicious spreadsheet formulas through a user-facing registration field, which are later exported and detonated when an operator opens the generated CSV, yielding code execution on the operator's workstation and potential privilege escalation. Tracked as CWE-1236 with a 9.8 CVSS and SSVC classifying it as PoC-available, automatable, with total technical impact; publicly available exploit code exists via a third-party GitHub write-up, but it is not on CISA KEV and EPSS is low at 0.10% (28th percentile). The vendor has published a security advisory and release notes covering the affected builds.
Remote code execution in Pyro (Python Remote Objects) 3.x arises because the wire protocol deserializes untrusted client messages with Python's pickle module, letting an attacker who can reach a Pyro daemon run arbitrary code by sending a crafted pickled message. All Pyro 3.x deployments that accept network connections are affected, and CWE-94 code injection combined with the CVSS 9.8 rating reflects full compromise of the host process. There is no public exploit identified at time of analysis, and the low EPSS score (0.02%) suggests no observed opportunistic exploitation despite the trivial technical requirements.
Command injection in KubePlus 4.14's kubeconfiggenerator component allows an attacker to run arbitrary shell commands on the host by supplying a crafted `chartName` value to the `/registercrd` endpoint. The flaw stems from unsanitized concatenation into a `subprocess.Popen(..., shell=True)` call, yielding full code execution (SSVC technical impact: total). A public proof-of-concept exists, but EPSS is very low (0.04%) and it is not on CISA KEV, so no public exploit indicates active exploitation at time of analysis.
SQL injection in Sourcecodester Online Thesis Archiving System v1.0's /otas/view_archive.php endpoint allows remote unauthenticated attackers to manipulate database queries, potentially extracting sensitive thesis data, authentication credentials, or modifying database contents. No public exploit identified at time of analysis, with minimal observed exploitation probability (EPSS 0.01%, 2nd percentile). The vulnerability affects a PHP-based academic archiving platform with limited deployment footprint.
Heap buffer overflow in FFmpeg 8.0.1's av_bprint_finalize() function enables remote denial-of-service attacks through maliciously crafted media files. Exploitation requires no authentication (CVSS AV:N/AC:L/PR:N/UI:N), making this accessible to any network-based attacker who can deliver manipulated input to vulnerable FFmpeg instances. EPSS score of 0.04% (12th percentile) indicates low observed exploitation probability, and no public exploit code or active exploitation has been identified at time of analysis. While CVSS rates this 7.5 HIGH due to availability impact, real-world risk is primarily limited to public-facing media processing services.
Resource deallocation flaw in FFmpeg 8.0.1's zmqsend.c utility enables remote denial of service through crafted input files. Attackers can trigger improper cleanup of allocated resources without authentication (CVSS AV:N/AC:L/PR:N/UI:N), exhausting system resources. EPSS score of 0.04% (11th percentile) indicates low real-world exploitation probability, and no active exploitation confirmed (not in CISA KEV). The vulnerability affects a non-core utility component used for ZeroMQ message sending, limiting practical attack surface compared to main FFmpeg libraries.
Out-of-bounds read in FFmpeg 8.0.1's AV1 decoder allows remote denial-of-service via malicious video files. Attackers craft inputs targeting read_global_param() in libavcodec/av1dec.c to trigger memory access violations, crashing the decoder. Affects applications processing untrusted AV1 video content (media servers, transcoders, browsers with FFmpeg). CVSS 7.5 (High) reflects network-exploitable DoS; EPSS 0.04% indicates low observed exploitation probability. No active exploitation confirmed (n
Totara LMS versions up to 19.1.5 allow unauthenticated remote attackers to conduct email bombing attacks by abusing the forgot password API endpoint, which lacks rate limiting on target email addresses. With a CVSS score of 9.8, the vulnerability enables complete compromise of confidentiality, integrity, and availability. Despite the critical score, EPSS estimates only 0.02% exploitation probability (5th percentile), and no active exploitation is confirmed (not in CISA KEV). A public proof-of-concept exists on GitHub, demonstrating the abuse technique.
Brute-force authentication bypass in Totara LMS versions 19.1.5 and earlier allows unauthenticated remote attackers to compromise user accounts via credential stuffing. The vulnerability chains login page manipulation with missing rate-limiting controls, enabling automated password guessing attacks. With CVSS 9.8 (critical) severity but only 0.02% EPSS probability (5th percentile), this represents a high-severity design flaw with currently low observed exploitation activity. A proof-of-concept exploit exists on GitHub (saykino/CVE-2026-31282), lowering the barrier for attack automation, though no confirmed active exploitation (CISA KEV) has been reported.
HTML Injection in Totara LMS through version 19.1.5 allows authenticated users with low privileges to inject malicious HTML/JavaScript into messages sent to all application users, enabling session hijacking and arbitrary command execution in victims' browsers. A publicly available exploit exists (GitHub POC referenced), though no confirmed active exploitation (not in CISA KEV). EPSS score of 0.02% indicates very low observed exploitation probability despite the CVSS 8.0 rating, suggesting limited attacker interest or opportunity in real-world environments.
Hardcoded debug credentials in BMC Control-M/MFT 9.0.20-9.0.22 enable unauthenticated remote attackers to access the MFT API debug interface with full system privileges. The vulnerability (CWE-798) stems from cleartext default credentials embedded in the application package, providing complete confidentiality, integrity, and availability compromise (CVSS 9.8). No public exploit identified at time of analysis, with EPSS score of 0.02% indicating low observed exploitation activity despite critical severity rating.
SQL injection in SourceCodester Engineers Online Portal v1.0 allows remote unauthenticated attackers to execute arbitrary SQL commands through the new_password parameter in update_password.php, enabling complete database compromise, authentication bypass, and potential server takeover. Despite a critical CVSS score of 9.8, EPSS indicates only 0.01% exploitation probability (2nd percentile), and no public exploit identified at time of analysis beyond technical documentation. The vulnerable PHP application lacks input validation on password update functionality, a common weakness in legacy web portals.
SQL injection in itsourcecode Online Student Enrollment System v1.0 allows unauthenticated remote attackers to execute arbitrary SQL commands via the 'subjcode' parameter in assignInstructorSubjects.php, potentially enabling complete database compromise, authentication bypass, and data exfiltration. CVSS 9.8 (Critical) reflects network-accessible exploitation requiring no privileges or user interaction. EPSS score of 0.01% (2nd percentile) indicates minimal observed exploitation activity, though publicly available exploit code exists via GitHub repository documentation. No vendor-released patch identified at time of analysis.
SQL injection in itsourcecode Online Student Enrollment System v1.0 allows unauthenticated remote attackers to extract, modify, or delete database contents via unsanitized classId parameter in instructorClasses.php. Despite critical CVSS 9.8 rating, EPSS score of 0.01% (2nd percentile) indicates minimal observed exploitation activity. Public proof-of-concept documentation exists on GitHub, lowering exploitation barrier for opportunistic attackers targeting educational institutions using this enrollment platform.
SQL injection in PHP-MYSQL-User-Login-System v1.0 allows unauthenticated remote attackers to execute arbitrary SQL commands via the username parameter at login.php, potentially enabling complete database compromise, authentication bypass, and unauthorized administrative access. Publicly available exploit code exists (GitHub POC). Despite a critical CVSS score of 9.8, EPSS exploitation probability is extremely low (0.01%, 2nd percentile), and no confirmed active exploitation (not in CISA KEV). The low real-world exploitation signal suggests limited deployment or attacker interest in this specific application.
SQL injection in itsourcecode Online Student Enrollment System v1.0 allows unauthenticated remote attackers to execute arbitrary SQL commands through the 'subjcode' parameter in scheduleSubList.php, achieving full database compromise with potential for complete system control. CVSS 9.8 (Critical) reflects network-accessible exploitation requiring no privileges or user interaction. EPSS score of 0.01% (2nd percentile) indicates minimal observed exploitation activity, and no CISA KEV listing exists, suggesting limited real-world targeting despite the theoretical severity. Public proof-of-concept documentation exists on GitHub, lowering the technical barrier for exploitation.
SQL injection in itsourcecode Online Student Enrollment System v1.0 allows unauthenticated remote attackers to execute arbitrary SQL commands via the 'coursename' parameter in newCourse.php, potentially enabling complete database compromise, authentication bypass, and server-side code execution. Despite a critical 9.8 CVSS score, EPSS indicates only 0.01% exploitation probability (2nd percentile), suggesting minimal observed attacker interest. Publicly available exploit code exists (GitHub POC), but no CISA KEV listing confirms active exploitation.
BMC Control-M/MFT 9.0.20-9.0.22 exposes API identifiers and secret values to unauthenticated remote attackers via an API management endpoint, enabling unauthorized invocation of privileged API operations. The CVSS vector (AV:N/AC:L/PR:N) confirms network-accessible, low-complexity exploitation requiring no authentication. Despite a 7.5 CVSS score, EPSS probability is minimal (0.03%, 8th percentile), and no public exploit or active exploitation (non-KEV) is identified at time of analysis.
SQL injection in BMC Control-M/MFT 9.0.20-9.0.22 debug API enables authenticated attackers to execute arbitrary SQL queries, leading to file system access and potential remote code execution. CVSS 8.8 reflects network-accessible, low-complexity exploitation requiring low-privilege authentication. EPSS score of 0.04% (13th percentile) indicates minimal current exploitation probability, and no public exploit identified at time of analysis. Patch PAAFP-9-0-22-025 addresses the vulnerability.
PowerStrip driver (pstrip64.sys) through version 3.90.736 enables authenticated local users to escalate privileges to SYSTEM by sending malicious IOCTL requests that map arbitrary physical memory into user-mode address space, allowing modification of kernel structures. EPSS score of 0.02% (5th percentile) indicates low automated exploitation likelihood, though publicly available exploit code exists (PacketStorm Security reference), making this a realistic threat in environments where PowerStrip is deployed and local access is possible. No CISA KEV listing indicates no confirmed widespread exploitation at time of analysis.
Command injection in ToToLink A3300R router firmware v17.0.0cu.557_B20221024 enables unauthenticated remote attackers to execute arbitrary OS commands via the stun-pass parameter in /cgi-bin/cstecgi.cgi endpoint. No public exploit identified at time of analysis, despite GitHub vulnerability disclosure repository. CVSS 9.8 (Critical) reflects trivial network-based exploitation without authentication, though EPSS probability remains low (0.01%, 2nd percentile), suggesting limited current attacker interest in this specific router model despite maximum severity rating.
DLL injection in GatewayGeo MapServer for Windows version 5 enables authenticated local attackers to escalate privileges to SYSTEM level through crafted executable placement. The vulnerability exploits insecure library loading paths, allowing low-privileged users to inject malicious DLLs that execute with elevated permissions. Publicly available exploit code exists. Affects Windows deployments only; CVSS 8.8 reflects local attack vector requiring low privileges but achieving full system compromise across security boundaries.
Remote code execution in OSGeo MapServer versions prior to 8.0 enables unauthenticated attackers to execute arbitrary code through dynamic-link library (DLL) injection via a specially crafted executable. The vulnerability requires no user interaction and has low attack complexity (CVSS 9.1 Critical), though real-world exploitation probability remains low (EPSS 2%, 5th percentile). Publicly available exploit code exists in a researcher's GitHub repository, but no confirmed active exploitation (CISA KEV) has been documented at time of analysis.
Server-Side Request Forgery in ERPNext 16.0.1 and Frappe Framework 16.1.1 enables unauthenticated attackers to force servers to make arbitrary HTTP requests to internal services through insufficiently sanitized HTML in Print Format PDF generation. Attackers inject HTML elements like <iframe> referencing external resources, which the PDF rendering engine automatically fetches server-side, exposing cloud metadata endpoints and internal network resources. No public exploit identified at time of analysis. CVSS 9.1 severity reflects network-accessible attack vector requiring no authentication or user interaction.
Buffer overflow in OpenAirInterface 2.2.0 AUSF component crashes service when processing oversized NAS PDU Authentication Response via UplinkNASTransport messages. Unauthenticated remote attackers can send malformed authentication responses (e.g., 100-byte payloads exceeding expected bounds) triggering AUSF component crash, preventing legitimate user registration and verification. Affects 5G core network deployments using OpenAirInterface AUSF. No public exploit identified at time of analysis. CVSS 7.5 High severity due to network-accessible denial of service without authentication requirements.
Integrity protection bypass in OpenAirInterface v2.2.0 allows unauthenticated network attackers to downgrade 5G security context by forcing acceptance of IA0-only capability during initial UE registration, despite NIA1/NIA2 being configured. Exploitation enables replay attacks against mobile network infrastructure through manipulation of Security Mode Complete messages, compromising session integrity without confidentiality impact. No public exploit identified at time of analysis.
Unauthenticated super administrator account creation in MRCMS 3.1.2 allows remote attackers to bypass all access controls and add privileged accounts directly via UserController.save() method. The vulnerability exposes full system compromise through network-accessible endpoints requiring no prior authentication. CVSS 9.8 critical severity reflects unrestricted administrative takeover. No public exploit identified at time of analysis; low observed exploitation activity (EPSS <1%).
Unauthenticated account creation bypass in megagao production_ssm v1.0 allows remote attackers to create super administrator accounts via direct API access to /user/insert endpoint. The UserController.java insert() method processes account creation requests without authentication enforcement (CVSS vector PR:N confirms unauthenticated access). Successful exploitation grants full administrative control, enabling attackers to compromise confidentiality, integrity, and availability of the entire application. No public exploit identified at time of analysis.
Authenticated remote code execution in Daylight Studio FuelCMS version 1.5.2 allows low-privileged users to execute arbitrary code via the Blocks module. CVSS 8.8 rating indicates network-accessible attack requiring low-complexity exploitation without user interaction, enabling full system compromise (confidentiality, integrity, availability impact). No public exploit identified at time of analysis. Low observed exploitation activity (EPSS 0.02%).
Authentication bypass in OpenAirInterface V2.2.0 Access Management Function (AMF) allows unauthenticated remote attackers to register unauthorized User Equipment (UE) devices on 5G core networks. Exploiting incorrect state machine transitions during UE registration, attackers send SecurityModeComplete messages after InitialUERegistration to trigger registration acceptance without completing proper authentication procedures. This grants full network access to malicious devices, enabling unauthorized subscriber services consumption, interception of traffic, and potential lateral movement within 5G infrastructure. No public exploit identified at time of analysis.
Authentication bypass in Kaleris Yard Management System (YMS) v7.2.2.1 enables unauthenticated remote attackers to completely circumvent login verification and gain unauthorized access to application resources with full confidentiality, integrity, and availability impact. The vulnerability has a 9.8 CVSS score with network-based attack vector requiring no privileges or user interaction. Currently tracked at 2% EPSS (5th percentile) with no confirmed active exploitation (not in CISA KEV), though a public proof-of-concept repository exists on GitHub, significantly elevating exploitation risk for this critical authentication flaw.
AMF daemon in OpenAirInterface V2.2.0 crashes upon receipt of malformed NGAP control-plane messages containing mismatched procedure codes or PDU type violations (e.g., successfulOutcome where InitiatingMessage is required). Remote attackers can trigger denial of service without authentication (CVSS AV:N/PR:N), exploiting improper input validation (CWE-20) in 5G core network signaling. Publicly available exploit code exists, SSVC framework classifies as automatable with partial technical impact, and EPSS data not provided but attack simplicity (AC:L) indicates low barrier to exploitation.
Percona PMM (Percona Monitoring and Management) versions prior to 3.7 allow authenticated users with pmm-admin privileges to execute arbitrary operating system commands by exploiting excessive database superuser privileges through the 'Add data source' feature. This privilege escalation vulnerability enables container/system breakout from database context to shell access. EPSS score is low (0.04%, 13th percentile) indicating minimal observed exploitation activity, and CISA SSVC assessment confirms no active exploitation with non-automatable attack characteristics, though technical impact is rated total.
Out-of-bounds read in Mbed TLS 3.x before 3.6.6 allows attackers to leak adjacent CCM context data through the multipart CCM API by passing an oversized tag_len parameter to mbedtls_ccm_finish(), which lacks validation against the internal 16-byte authentication buffer. Mbed TLS 4.x contains the same vulnerability in internal code but does not expose the vulnerable function publicly; exploitation requires direct application-level invocation of the affected API. No public exploit code or active exploitation has been reported, but the attack requires no special privileges.
TOCTOU race condition in Balena Etcher for Windows (versions prior to 2.1.4) enables local privilege escalation to arbitrary code execution when attackers replace legitimate scripts with malicious payloads during disk flashing operations. The vulnerability requires low privileges and user interaction but achieves high impact across confidentiality, integrity, and availability with scope change. No public exploit identified at time of analysis, though technical details are available via researcher disclosure (B1tBreaker). EPSS data not available, but the local attack vector and high complexity reduce immediate remote exploitation risk.
Mbed TLS versions 3.5.0 through 4.0.0 allow client impersonation during TLS 1.3 session resumption, enabling an attacker to assume the identity of a legitimate client when reestablishing a previously negotiated session. The vulnerability affects the session resumption mechanism in TLS 1.3 and permits information disclosure; no CVSS score or exploit status data is currently available from public sources.
Finite-field Diffie-Hellman (FFDH) in Mbed TLS 3.5.x, 3.6.0 through 3.6.5, and TF-PSA-Crypto 1.0 lacks contributory behavior due to improper validation of peer-supplied parameters, allowing an attacker to restrict the shared secret to a small set of predictable values. While the vulnerability does not directly impact TLS (which does not depend on contributory behavior), it poses a significant risk to protocols that do rely on this property, including those where an active network attacker or malicious peer can exploit the weakness. No CVSS score or public exploit code has been assigned at the time of analysis.
NULL pointer dereference in Mbed TLS distinguished name (X.509) parsing allows remote attackers to trigger a denial of service by writing to address 0, affecting Mbed TLS versions 3.6.5 and earlier, and 4.0.0. The vulnerability is reachable during X.509 certificate processing and does not require authentication. No public exploit code or active exploitation has been confirmed at the time of analysis.
Mbed TLS before version 3.6.6 and TF-PSA-Crypto before version 1.1.0 contain a PRNG seed misuse vulnerability that enables information disclosure. An attacker who gains access to a seeded PRNG instance can potentially predict or replicate pseudo-random number generation, compromising cryptographic material confidentiality. The vulnerability affects cryptographic libraries used in embedded systems and IoT devices, with confirmed availability of vendor security advisories but no CVSS score assigned at time of analysis.
Buffer overflow in Mbed TLS versions 3.5.0 through 3.6.5 allows remote attackers to cause a denial of service or potentially execute arbitrary code via crafted input to the x509_inet_pton_ipv6() function used in X.509 certificate parsing. The vulnerability is fixed in Mbed TLS 3.6.6 and 4.1.0. No public exploit code or confirmed active exploitation has been identified at the time of analysis.
Buffer overflow in Mbed TLS public key export functionality for Finite Field Diffie-Hellman (FFDH) keys affects versions through 3.6.5 and TF-PSA-Crypto 1.0.0. An attacker can trigger a memory corruption condition during FFDH public key export operations, potentially enabling code execution or denial of service depending on memory layout and application context. No public exploit code or active exploitation has been confirmed at time of analysis.
Remote code execution in DedeCMS 5.7.118 allows unauthenticated attackers to execute arbitrary code through crafted setup tag values during module upload operations. The vulnerability exploits insufficient input validation in the module upload functionality, enabling direct code injection. No CVSS score, EPSS data, or KEV confirmation is available; however, the presence of a public proof-of-concept demonstrates practical exploitability.
SQL injection in pandas-ai v3.0.0 allows remote code execution through the pandasai.agent.base._execute_sql_query component, enabling attackers to manipulate SQL queries and potentially access, modify, or exfiltrate database contents. No CVSS score, EPSS data, or KEV status is available; however, the vulnerability affects a widely-used data analysis library and publicly available proof-of-concept code exists, elevating real-world risk despite incomplete severity metrics.
Buffer overflow in TOTOlink A3600R v5.9c.4959 setAppEasyWizardConfig interface allows remote code execution or denial of service via unvalidated rootSsid parameter in /lib/cste_modules/app.so. The vulnerability affects a Wi-Fi router's configuration endpoint and enables unauthenticated attackers to trigger memory corruption with potential for arbitrary code execution. No CVSS vector or patch status was available at time of analysis.
Pharmacy Product Management System 1.0 accepts negative price and total cost values in sales transactions due to insufficient input validation in add-sales.php, enabling attackers to manipulate financial records, corrupt sales reports, and cause financial loss. The vulnerability allows unauthenticated or low-privilege users to submit arbitrary negative values that bypass business logic controls. Publicly available exploit code exists demonstrating this business logic flaw.
Arbitrary file overwrite in Docudepot PDF Reader v1.0.34 enables attackers to overwrite critical internal files through the file import process, resulting in arbitrary code execution or information exposure. The vulnerability affects the mobile PDF viewer application across Android platforms. No public exploit code or active exploitation has been confirmed at time of analysis, though the severity of potential impact (RCE) warrants immediate investigation and patching.
Arbitrary file overwrite in Ora Tools PDF Reader & Editor APP v4.3.5 enables attackers to overwrite critical internal files through the file import process, resulting in arbitrary code execution or information exposure. The vulnerability affects the Android application and has been publicly disclosed; however, CVSS scoring, CISA KEV status, and vendor patch availability have not been independently confirmed at time of analysis.
Arbitrary file overwrite in Zora: Post, Trade, Earn Crypto v2.60.0 enables attackers to overwrite critical internal files through the file import process, resulting in remote code execution or information exposure. The vulnerability affects the cryptocurrency trading application's file handling mechanism, allowing unauthenticated remote attackers to inject malicious content into system-critical files. No active exploitation has been confirmed at time of analysis, though the attack vector and impact severity warrant immediate investigation by affected users.
Arbitrary file overwrite in InTouch Contacts & Caller ID APP v6.38.1 allows remote attackers to overwrite critical internal files through the file import process, enabling arbitrary code execution or sensitive information exposure. Affected versions are limited to 6.38.1; no CVSS score, EPSS, or active exploitation status (KEV) is available at this time, though the vulnerability chain to RCE presents material risk.
Arbitrary file overwrite in PEAKSEL D.O.O. NIS Animal Sounds and Ringtones v1.3.0 allows attackers to overwrite critical internal files during the file import process, enabling remote code execution or sensitive information exposure. The vulnerability affects the application's import functionality without requiring authentication. No public exploit code or active exploitation has been independently confirmed at the time of analysis.
Arbitrary file overwrite in My Location Travel Timeline v11.80 by Squareapps LLC permits attackers to overwrite critical internal files through the file import process, resulting in arbitrary code execution or information disclosure. Attack vector and complexity details are not confirmed from available CVSS data, and active exploitation status is unconfirmed.
Arbitrary file overwrite in Funambol Zefiro Cloud v32.0.2026011614 allows attackers to overwrite critical internal files during the file import process, enabling remote code execution or information disclosure. The vulnerability affects the cloud application and its associated mobile client. No CVSS score or official vendor patch has been assigned as of analysis time, though the reported impact (RCE/information exposure) is severe.
Arbitrary file overwrite in UXGROUP LLC Cast to TV Screen Mirroring v2.2.77 enables remote attackers to overwrite critical application files through a malicious file import process, resulting in remote code execution or information disclosure. No CVSS score, exploit code availability, or active exploitation status confirmed from available data.
Arbitrary file overwrite in FLY is FUN Aviation Navigation v35.33 permits attackers to overwrite critical internal files through the file import process, enabling remote code execution or information disclosure. No CVSS score, CVE severity classification, or patch status has been established. The vulnerability affects a niche aviation navigation software product with limited public disclosure.
Arbitrary file overwrite in PDF Reader App TA/UTAX Mobile Print v3.7.2.251001 allows remote attackers to overwrite critical internal files during the file import process, potentially leading to remote code execution or unauthorized information exposure. The vulnerability affects a mobile print utility with demonstrated proof-of-concept documentation available on GitHub, though CVSS scoring and formal vendor patch status remain unavailable at time of analysis.
Arbitrary file overwrite in MaruNuri LLC v2.0.23 allows remote attackers to overwrite critical internal files during the file import process, enabling arbitrary code execution or information exposure. No CVSS score, exploit code availability, or active exploitation status is documented in available sources.
Arbitrary file overwrite in DeftPDF Document Translator v54.0 permits attackers to overwrite critical internal files through the file import mechanism, potentially enabling remote code execution or sensitive information exposure. The vulnerability affects DeftPDF Document Translator specifically at version 54.0 and is documented by academic researchers at Fudan University's security systems group. Attack complexity and authentication requirements cannot be definitively assessed due to missing CVSS vector data, though the file import process suggests user interaction may be required.
Remote code execution in Ridvay Code's command auto-approval module allows unauthenticated attackers to bypass whitelist security controls via shell command substitution syntax (e.g., $(...) or backticks) embedded in command arguments. The vulnerability stems from insufficient regular expression validation that fails to detect command injection payloads, permitting an attacker to execute arbitrary OS commands with automatic approval. No user interaction is required; a crafted command such as git log --grep="$(malicious_command)" will be misidentified as safe and executed by the underlying shell, resulting in remote code execution.
Directory traversal in agentic-context-engine up to version 0.7.1 enables arbitrary file writes through the checkpoint_dir parameter in OfflineACE.run, exploiting inadequate path normalization in the save_to_file method. Unauthenticated attackers can overwrite arbitrary files within the application process's permissions scope, potentially achieving code execution, privilege escalation, or application compromise depending on deployment context and file system layout.
Remote code execution in DSAI-Cline's command auto-approval module allows unauthenticated attackers to bypass whitelist validation by embedding literal newline characters within command payloads, forcing the system to execute arbitrary OS commands without user interaction. The vulnerability exploits ineffective string-based parsing that fails to sanitize newline separators, enabling attackers to chain whitelisted commands (e.g., git log) with malicious code that PowerShell interprets as sequential commands. No CVSS score, EPSS data, or KEV confirmation available; exploitation status and real-world impact remain unconfirmed.
Remote code execution in Ridvay Code's command auto-approval module allows unauthenticated attackers to bypass whitelist protections via shell command substitution syntax ($(…) and backticks) embedded in seemingly benign git commands, achieving code execution without user interaction. The vulnerability exploits inadequate regular expression validation that fails to detect shell metacharacters in command arguments, enabling attackers to inject arbitrary commands that execute with the privileges of the Ridvay Code process.
InfCode's terminal auto-execution module fails to properly validate PowerShell commands due to an ineffective blacklist and lack of semantic parsing, allowing attackers to bypass command filtering through syntax obfuscation. When a user imports a specially crafted file into the IDE, the Agent executes arbitrary PowerShell commands without user confirmation, leading to remote code execution or sensitive data exfiltration. No public exploit code or active exploitation has been confirmed at time of analysis.
Prompt injection attacks in Sixth's automatic terminal command execution feature bypass the model-based safety classification system, allowing attackers to execute arbitrary commands without user approval by wrapping malicious payloads in templates that mislead the AI into treating them as safe operations.
Remote code execution in DSAI-Cline's command auto-approval module allows unauthenticated attackers to bypass whitelist validation by embedding newline characters in command payloads, forcing automatic approval and sequential execution of arbitrary OS commands via PowerShell without user interaction.
HAI Build Code Generator's automatic command execution feature can be bypassed through prompt injection attacks, allowing unauthenticated remote code execution by misleading the AI model into misclassifying malicious commands as safe. The vulnerability exploits a fundamental design flaw in the model's safety classification logic, where attackers can wrap destructive commands in generic templates to bypass the user approval requirement that should be triggered for potentially dangerous operations.
SakaDev's automatic terminal command execution feature can be bypassed via prompt injection attacks, allowing unauthenticated remote attackers to execute arbitrary commands without user approval by wrapping malicious commands in templates that mislead the underlying language model into misclassifying destructive operations as safe. The vulnerability exploits a design flaw in the model-based safety classification mechanism rather than a traditional code defect, affecting the extension across all versions where the 'Execute safe commands' option is enabled.
Remote code execution in Syntx's command auto-approval module allows unauthenticated attackers to bypass whitelist security via shell command substitution syntax in command arguments. The vulnerability exploits inadequate regular expression parsing that fails to detect $(…) and backtick command substitution patterns, enabling an attacker to inject malicious commands within seemingly benign git operations (e.g., git log --grep="$(malicious_command)") that are automatically approved and executed with full system privileges. No CVSS score or KEV status data available; no public exploit code confirmed at time of analysis.